(Almost) Unsupervised Grammatical Error Correction using a Synthetic Comparable Corpus

摘要

We introduce unsupervised techniques based on phrase-based statistical machine translation for grammatical error correction (GEC) trained on a pseudo learner corpus created by Google Translation. We verified our GEC system through experiments on a low resource track of the shared task at Building Educational Applications 2019 (BEA2019). As a result, we achieved an F_(0.5) score of 28.31 points with the test data.